A Tsinghua University doctoral student dropped out in 1998 to build robots and raised nearly 500 million yuan in a month.

ChinaVenture Capital Exclusive has learned that Beijing-based humanoid robot company Songyan Dynamics has completed a nearly 200 million yuan Pre-B+ round of financing. This round of financing was led by CICC Capital, with Yuntai Capital and Houwei Capital participating as follow-on investors. The financing will be used to increase investment in technological innovation and R & D, expand ecological cooperation in high-value application scenarios, and establish a path from R & D to large-scale mass production and delivery.

Humanoid robots and embodied intelligence are the most capital-intensive sectors this year. At least judging from the frequency and rounds of financing, this is a company on the rise. After completing a nearly 300 million yuan Pre-B round of financing led by Fangguang Capital on October 26th, it has now received an additional nearly 200 million yuan in support. The interval between the two announced financings is less than a month, with a total financing amount of nearly 500 million yuan, and a total of five rounds of financing have been completed within the year.

Also announced was a strategic cooperation with "Programming Cat": the two parties will jointly build a "Humanoid Robot Programming Education Laboratory". Last month, the company just launched its first high-performance humanoid robot "Bumi Xiaobumi" priced under 10,000 yuan, mainly targeting technology enthusiasts and teenagers learning programming. While the industry is generally exploring B2B scenarios, these two developments connect Songyan's next plan: using the education scenario as an entry point to penetrate the quasi-consumer market.

The guest invited in this issue of "Blue Hour" is Jiang Zheyuan, the founder of Songyan Dynamics. At the age of 25, he dropped out of his Ph.D. program at Tsinghua University and was determined to develop humanoid robots. Although his identity is gradually restricting his expression, at least for now, he is a 1998-born entrepreneur who still retains a sense of vitality. When facing each other, he doesn't have strict defenses and excessive concerns. During casual conversations, he will candidly admit his personal weaknesses and reflect on his lack of business knowledge. What he talks about the most is that in the past, he only knew to focus on technology and ignored building connections with the market.

This is quite representative. Most entrepreneurs who come from universities have to go through this stage. These people riding on the wave of the times are obsessed with technology and are used to thinking from a scientific perspective - just like when he set the price of the robot at 9,998 yuan, he used the "extreme assumption method" taught by his father when he was doing Olympiad math problems. In the process of entering the market, even professors who are more favored by capital will inevitably encounter difficulties in realizing and mastering another set of communication and rules. They have to experience some pain before they can figure it out. I once heard that a popular embodied intelligence founder also spent some time to figure out how to tell stories with demos before opening up the situation in financing.

In this regard, youth may not necessarily be a disadvantage. At least, young people are more flexible. After experiencing a period of "being unrecognized for one's talents", the founder has books such as "Learn to Speak", "Influence", and "Documentaries Also Need to Tell Stories" recommended by investors on his bookshelf. His awareness of external expression has awakened.

Of course, the challenges are not limited to this. "In the past two years of entrepreneurship, I have experienced more things than in the previous twenty years. I'm being tempered every day," Jiang Zheyuan frowned. He has recruited a group of senior executives older than him and calls most of them "brother" and "sister" in the company. He is still working hard to learn what true "management" is. After winning the second place in the half - marathon race for humanoid robots in April, the company's momentum has rapidly gathered, and the proposition has changed from how to "survive" to how to "stay in the game".

Last week, in the office located in the Changping Future Science City, we had a chat with Jiang Zheyuan about the latest products and strategies, and also traced the original intention of entrepreneurship and the necessary experiences. He brewed a pot of white orchid tea produced in Ya'an, Sichuan. Even though he doesn't like tea himself, setting up a tea set and brewing a pot of steaming water to make tea shows his hospitality as the leader of the company. Outside the window is an open space where plants grow wildly in spring and summer, just like this company that has been established for only a little over two years.

If I had to use a color to describe him, I think the color of young trees would be the most appropriate. The bright yellow - green color is full of the vitality of spring - you can't plan its growth, but you can feel an unadorned vitality.

Many people say we are engaging in a price war. This idea is completely wrong.

Liu Yanqiu: How did you come up with the idea of making a humanoid robot priced under 10,000 yuan?

Jiang Zheyuan: Look at any industry. In the early stage, products are always expensive, have low performance, and are produced in small quantities. Later, they gradually become cheaper, and now consumers can afford them. Any industry will go through a stage of low production volume, high gross profit, and high prices in the early stage. The ultimate outcome must be large - scale production, reasonable gross profit, and affordable prices. Just like today's cars, the gross profit is about 20%, and some are even lower, but the production volume is extremely large, and the prices are acceptable to the public.

I firmly believe that humanoid robots will eventually become popular and enter every family. To be honest, it's quite difficult for robots to really enter households. They are expensive and can't really solve practical problems for people. Thinking about it this way, it seems like an insoluble problem. But I have a thinking habit taught by my father. When he taught me to do Olympiad math problems, he used the "extreme assumption method": if you can't solve a problem, set one of the numbers to an extreme value, such as 0 or infinity, and then see if the problem becomes easier to solve. I applied this method to robots: since it's difficult for robots to enter households, what if we take the "price" factor to the extreme? If humanoid robots are extremely cheap, is it possible to achieve large - scale production? Following this line of thought, if we can make a humanoid robot priced under 10,000 yuan, even if its current functions are not as good as expected, once the price threshold is broken, it will most likely be the first to enter consumers' homes.

Price is just a means, not an end. We are not making products cheap just for the sake of being cheap. The core is to make robots have scale and become popular. The common goal in the industry should be to promote the popularization of robot technology. This is extremely important.

Many people say we are engaging in a price war. This idea is completely wrong. Previously, Unitree's R1 robot was sold for 39,900 yuan. Some companies set their prices slightly lower than Unitree's, and some even offered limited - time discounts. This is a price competition, which is essentially trying to grab market share from Unitree. But we have brought the price down from tens of thousands of yuan to a few thousand yuan. This is not competition: the products priced at 39,900 yuan and slightly lower target the same group of customers. In essence, it's about using low prices to cut a larger share of the existing market, which is an in - house competition. But when we bring the price down to around 9,000 yuan, we are expanding the market. The customers for products priced at 39,900 yuan are the original group, while the customer base for products priced at 9,000 yuan will be much larger. I'm looking for my own share in a huge new customer pool. This is not competition; it's market expansion.

Liu Yanqiu: When selling the product at 9,998 yuan, what's the approximate cost?

Jiang Zheyuan: I can't disclose the specific figure, but we can definitely make a profit.

Liu Yanqiu: Why is it your company that is doing this? What methods are used to reduce costs? What have you sacrificed in terms of technology and product to reduce costs?

Jiang Zheyuan: I think a company needs to find its correct position in the industry. In the market for robots priced at tens of thousands of yuan, it's very difficult to create significant product differentiation no matter how you define the product. Unitree has done the best in this field. So we want to change our approach and look for opportunities in a broader scope.

As for why we can make robots at such a low cost, the underlying logic is simple: 80% of the cost of consumer electronics products is determined by the design, and 20% is determined by the production volume. Whether the cost can be optimized well is absolutely related to whether the boss attaches importance to it. If cost optimization is a top - priority project, it will definitely be done well.

For example, we require structural engineers to design products specifically for mold - making during the design stage, and details such as draft angles will be handled very well. In terms of material selection, we ask engineers to design robots according to the characteristics of composite materials. Because the overall cost of composite materials is much lower than that of metal materials and the multi - step die - casting process. In addition, we have promoted vertical integration of the supply chain. We can achieve this because on the one hand, we have developed several generations of robots and have accumulated strong design capabilities, and on the other hand, it also reflects our supply chain management capabilities.

Liu Yanqiu: Can you be more specific about when you started to conceive the strategy of exploring the consumer - grade robot market with "Xiaobumi"?

Jiang Zheyuan: This can't really be called a strategy. It's just a short - term approach for us. From the time we really made up our minds to develop Xiaobumi to the time we launched the product, the time span wasn't very long. Our company has always had strong execution ability and high efficiency. Many people praise me for my strong execution ability, which is what many people think is great about me. But I think it's a simple thing. The core is "once you have an idea and think it's right, just do it". It sounds simple, but many people just can't do it.

Liu Yanqiu: So in your opinion, reducing costs isn't that difficult?

Jiang Zheyuan: I think so.

Liu Yanqiu: You said this is a strategy of differential competition. So what's the current price ecosystem in the robot industry? Where does your company stand in it?

Jiang Zheyuan: We definitely have the lowest price in the same category. Unless compared with toy - level products that can't be really used, among robots with real practical value, our price is the lowest.

Liu Yanqiu: Don't you also have products priced at over 30,000 yuan?

Jiang Zheyuan: That was the limited - time pre - sale price before May Day for the E1 and N2 robots. After May Day, the price has returned to normal. Now, there are different versions. The cheapest one is 69,900 yuan, and there is also an EDU version priced at over 100,000 yuan, targeting high - end scientific research scenarios.

Liu Yanqiu: So the scenarios covered by your product line mainly include scientific research, performances, and the consumer market?

Jiang Zheyuan: Both performances and scientific research are included. The Xiaobumi product actually straddles both the B2B and B2C markets: in the B2B market, it targets educational institutions and event partnerships; in the B2C market, it clearly targets family consumers with children aged 7 - 13.

Liu Yanqiu: Currently, most humanoid robots are purchased by B2B customers, such as pharmacies, factories, logistics companies, and scientific research institutions. Some are even provided with a trial - and - error scenario by industrial shareholders. How did you judge that the time to sell to C - end users had come?

Jiang Zheyuan: This judgment is mainly based on our own capabilities, rather than simply looking at the industry. We felt that we had the ability to do this at that time: we had a certain order base, had in - depth cooperation with supply chain partners, and had the ability to negotiate prices with the supply chain. When the numbers added up, we took action. In fact, from an industry perspective, there was a possibility of cost - reduction and low - price product development in 2024, but at that time, we didn't have the ability, and the timing wasn't right for us. Now that we have the ability, we have resolutely advanced. More established companies in the industry actually had the qualification to do this a long time ago.

Liu Yanqiu: Since those other companies had the ability, why didn't they explore this direction at that time?

Jiang Zheyuan: Maybe they didn't need to. At that time, the entire market was still a blue ocean, and there were clear sales channels for products in the higher - price segment. There was no need to do something unproven and risky. After all, no one could be sure whether reducing the price to such a low level would be successful.

Liu Yanqiu: So the main reason for your company to choose the consumer market now is that the original blue - ocean scenarios have become highly competitive?

Jiang Zheyuan: Yes, we are trying to avoid competition. In fields such as scientific research, education, and commercial performances, commercial performances aren't particularly competitive, but I think the industry ceiling is low, and it's not a long - term, large - scale development track.

Liu Yanqiu: I want to know if your cost - reduction methods are non - replicable. Can other companies easily catch up?

Jiang Zheyuan: They are not non - replicable. Any company with sufficient technical strength can achieve this. The real barrier in an industry actually comes from the flywheel effect. The better the business, the better it will become, forming a positive cycle. For example, in the automobile manufacturing industry, the logic is: the more cars are sold, the lower the supply chain cost will be, and the product price can be further reduced, which in turn allows the company to capture a larger market share, and then promotes cost - reduction and sales growth. This is the most typical Matthew effect. Our core idea in developing low - price consumer - grade humanoid robots is to be an early mover by taking advantage of the time difference in perception and quickly build the Matthew effect so that we can stand firm in the market.

Now we have launched products in this price range earlier than others. Later, other companies may also develop similar products, but before they catch up, we can first occupy ecological positions, such as locking in marketing resources, building sales channels, and enhancing brand power, and then further reduce costs through large - scale production. Once these things are done well, the Matthew effect will be formed: the more products are sold, the lower the cost will be, the more brand exposure there will be, and the sales will increase further. When we sell 20,000 units, the products will be everywhere in daily life, which will bring more natural exposure and further boost sales.

Liu Yanqiu: Are there any competitors in the industry that you admire?

Jiang Zheyuan: I really admire Unitree. They have persisted for many years and patiently waited for the industry to boom. This is really remarkable.

Liu Yanqiu: Did you benchmark against Unitree when you first started your business? They also just started developing humanoid robots at that time.

Jiang Zheyuan: We didn't deliberately benchmark against them, but we were indeed quite concerned about their actions.

Liu Yanqiu: So is the difference between Songyan and Unitree also reflected in different application scenarios?

Jiang Zheyuan: The scenarios are completely different. Unitree mainly focuses on the scientific research field.

Liu Yanqiu: They have occupied the ecological niche in the scientific research scenario.

Jiang Zheyuan: Yes, I've always thought that if you can't be at the top in a specific niche market, it's not worth doing.

Liu Yanqiu: Then can Unitree or other companies also enter the market at this price level if they want to?

Jiang Zheyuan: This actually involves a common phenomenon called "the innovator's dilemma". In simple terms, if you are already at the top of a small hill, you probably won't want to come down and climb another potentially taller mountain, even if the peak of that mountain is higher. Just like in the shampoo industry before, a leading company developed silicone - free shampoo by itself. Logically, it's better than silicone - based shampoo, but they didn't dare to launch it. If they did, it would impact their previously well - selling silicone - based shampoo, and it wouldn't make sense in terms of marketing. This is the innovator's dilemma. In the end, someone else made it successful. Few people are willing to give up their current stability and bet on an uncertain future.

Liu Yanqiu: Aren't customers who buy expensive products and those who buy cheap products two different groups